MathGroup Archive 2005

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Why is y a local variable here?

  • To: mathgroup at
  • Subject: [mg63189] Re: Why is y a local variable here?
  • From: Peter Pein <petsie at>
  • Date: Sat, 17 Dec 2005 03:46:21 -0500 (EST)
  • References: <> <> <dnv444$oas$>
  • Sender: owner-wri-mathgroup at

Steven T. Hatton schrieb:
> On Thursday 15 December 2005 11:27, Omega Consulting wrote:
>>I have a hard time understanding the purpose of your example. You say
>>that you want to define
>>sd[y_] = SDP[n - 1, y]
>>in your function. But you want the global value of y to be used. If
>>so, why have y as a parameter of sd. If the global value comes in,
>>then sd will always return the same thing (n and y are fixed values).
>>If you want the function to use the global value, don't make it a
>>parameter of sd.
> I didn't want Global`y to be used, I just didn't understand why it was being 
> hidden (renamed) when code using almost the same construct was seeing the 
> Global`y.
>>The other part of your question is where does y$ come from, why isn't
>>it y (which would then evaluate into the global y value). This is a
>>subtle side-effect of how Mathematica scoping works. It is somewhat
>>complicated and usually only appears when you nest one scoping
>>construct inside another. In this case, the = for sd inside the :=
>>for SDP. The details are in section 2.7.4 of the Mathematica Book
>>(Advanced Topic: Variables in Pure Functions and Rules). That
>>describes the situation in a much better way than I could reproduce
> That was my only question. Thank you.  I will need to read that section 
> carefully a few times. I believe the replacement is due to there being an 
> unbound variable "n" within the sd[y_]=SDP[n-1,y];, and that variable is 
> changed during evaluation.  The discussion seems to focus on pure functions, 
> but I suspect the same rules apply to Modules and other block structures as 
> well.  That seems to be what the rather cryptic comment about lambda 
> expressions is about.
> Here is another nuance of the situation.  The following is the correct 
> implementation of the algorithm.  If I change it by making y local to the 
> Module, it does not us "x" in the result.  Instead, the result contains the 
> instance variable of the Module.
> SDP[0, x_] := x
> SDP[n_Integer?Positive, x_] := Module[{sd, srp = Sqrt[Prime[n]], py},
>     sd[y_] = SDP[n - 1, y];
>     Expand[sd[x + srp] sd[x - srp]]]
> SDP[5, x]
> Maeder makes in interesting observation at the bottom of page 221 in 
> _Programming in Mathematica_.
> Steven
Hi Steven,

  if you want to save evaluation time with the sd[]-construct, this will fail, 
because sd[] will be redefined with evevery call of SDP[]. In that case you 
should follow the standard procedure:
SDP[0,x_]=x; SDP[n_,x_}:=SDP[n,x]=...

  • Prev by Date: Re: EUREKA Re: Types in Mathematica, a practical example
  • Next by Date: Re: EUREKA Re: Types in Mathematica, a practical example
  • Previous by thread: Re: Why is y a local variable here?
  • Next by thread: Re: Why is y a local variable here?